Journal of Vocational Behavior 66 (2005) 2644

www.elsevier.com/locate/jvb

Flow among music teachers and their

students: The crossover of peak experiencesq
Arnold B. Bakker*
Utrecht University, Department of Social and Organizational Psychology,
and Research School Psychology and Health, Utrecht, The Netherlands
Received 13 May 2003
Available online 4 February 2004

Abstract
This study among 178 music teachers and 605 students from 16 dierent music schools examined the peak experience of ow (Csikszentmihalyi, 1990). On the basis of the literature, it
is hypothesized that job resources, including autonomy, performance feedback, social support, and supervisory coaching have a positive inuence on the balance between teachers
challenges and skills, which, in turn, contributes to their experience of ow (absorption, work
enjoyment, and intrinsic work motivation). In addition, using emotional contagion theory, it
is hypothesized that ow may crossover from teachers to their students. The results of structural equation modeling analyses oer support for both hypotheses. These ndings are discussed in light of theories about crossover and emotional contagion.
2003 Elsevier Inc. All rights reserved.
Keywords: Job resources; Flow; Crossover; Emotional contagion

1. Introduction
Flow is a state of consciousness where people become totally immersed in an activity, and enjoy it intensely. According to Csikszentmihalyi (1997), such a peak experience can emerge in any situation in which there is activity, and researchers have
q
I thank Korine Scheeres for her help with data collection, and Joy Oliver for her valuable comments
on a previous draft of this paper.
*
Fax: +31-30-2537584.
E-mail address: a.bakker@fss.uu.nl.

0001-8791/$ - see front matter 2003 Elsevier Inc. All rights reserved.
doi:10.1016/j.jvb.2003.11.001

A.B. Bakker / Journal of Vocational Behavior 66 (2005) 2644

27

indeed found evidence for ow during the execution of a large number of dierent
activities, including sports, work, and playing music (Catley & Duda, 1997;
Csikszentmihalyi & Csikszentmihalyi, 1988; Csikszentmihalyi & LeFevre, 1989;
Jackson & Marsh, 1996; Kowal & Fortier, 1999). In the present study, the phenomenon of ow is investigated among music teachers and their students. The main research questions addressed whether job resources facilitate ow at work, and
whether this experience may crossover from music teachers to their students.

2. Flow at work
What is ow? The literature reveals a laundry list of denitions of the concept.
Csikszentmihalyi (1990), who coined the term ow, describes the concept as: The
state in which people are so intensely involved in an activity that nothing else seems
to matter; the experience itself is so enjoyable that people will do it even at great cost,
for the sheer sake of doing it. In addition to the pleasure in the activity and the intrinsic motivation to continue doing it, the total immersion in an activity seems to be
a central aspect of the ow-experience. For example, Lutz and Guiry (1994) describe
ow as: . . . a state of mind sometimes experienced by people who are deeply involved in some event, object, or activity. . . they are completely and totally immersed
in it. . . Indeed, time may seem to stand still and nothing else seems to matter while
engaged in the consumption event. These scholars emphasize the idea that time ies
during a ow experience. Ellis, Voelkl, and Morris (1994) dene ow as an optimal
experience that is the consequence of a situation in which challenges and skills are
equal. According to these researchers, such a situation facilitates the occurrence of
ow-related phenomena, such as positive aect, arousal, and intrinsic motivation
(p. 337). Furthermore, Ghani and Deshpande (1994) particularly emphasize the total
concentration and the enjoyment people experience when in ow: The two key
characteristics of ow are (a) total concentration in an activity and (b) the enjoyment
which one derives from an activity. . . There is an optimum level of challenge relative
to a certain skill level. (p. 383).
The most prominent denitions of ow seem to have three elements in common,
namely absorption (the total immersion in an activity), enjoyment, and intrinsic
motivation. These three elements are indeed the core components that are usually included in studies by many ow-researchers (e.g., Csikszentmihalyi, 1997;
Csikszentmihalyi & Csikszentmihalyi, 1988; Csikszentmihalyi, Rathunde, & Whalen,
1993; Larson & Richards, 1994). Accordingly, when ow is applied to the work situation, it can be dened as a short-term peak experience at work that is characterized by absorption, work enjoyment, and intrinsic work motivation. Absorption
refers to a state of total concentration, whereby employees are totally immersed in
their work. Time ies, and they forget everything else around them (cf.
Csikszentmihalyi, 1990). Employees who enjoy their work and feel happy make a
very positive judgment about the quality of their working life (cf. Veenhoven,
1984, 1996). This enjoyment or happiness is the outcome of cognitive and aective
evaluations of the ow experience (cf. Diener, 2000; Diener & Diener, 1996). Finally,

28

A.B. Bakker / Journal of Vocational Behavior 66 (2005) 2644

intrinsic work motivation refers to the need to perform a certain work-related activity
with the aim of experiencing the inherent pleasure and satisfaction in the activity (cf.
Deci & Ryan, 1985). Intrinsically motivated employees are continuously interested in
the work they are involved in (Harackiewicz & Elliot, 1998). Employees who are motivated by the intrinsic aspects of their work tasks want to continue their work; they
are fascinated by the tasks they perform (Csikszentmihalyi, 1997).

3. What are the antecedents of ow?

Csikszentmihalyi (1990) discovered that artists and athletes are particularly
prone to experiencing ow. Does this also apply to music teachers? According to
Klausmeier (1978), the practice, performance, and transference of music oers many
reasons to become totally immersed in the activity. The identication with the music
allows the musician to feel at one with and become absorbed in the music (Palmer,
1989). In addition, the high level of eort that is required to perform well coincides
with high concentration. We may therefore expect that musicians frequently experience ow. However, what are the factors in music teachers working environment
that contribute to ow experiences?
Csikszentmihalyis (1997) experience sampling studies have shown that people
more often experience ow during their work than during free time. These studies
show, for example, that boredom is more likely when people spend their leisure time
on activities such as watching television. This suggests that one has to invest time
and energy to experience ow. Indeed, research has shown that people need challenges to facilitate ow in a range of activities, including work. Moreover, researchers generally agree that the occurrence of ow is most likely when people perceive a
balance between the challenge of a situation and their own skills to deal with this
challenge (e.g., Clarke & Haworth, 1994; Csikszentmihalyi, 1990; Ellis et al., 1994;
Massimini & Carli, 1988). For example, a tennis player experiences a balance when
he is confronted with an opponent who is approximately equally skilled. In this situation, an exciting game may develop in which the tennis players have to do their
utmost to beat the opponent. On the contrary, theoretically, there is a higher probability of boredom if the opponent has fewer skills (low challenge in the situation,
self high skills), and a high probability of stress if the opponent is much better (high
challenge, low skills). There is indeed some empirical evidence for this pattern of experiences (e.g., Csikszentmihalyi, 1997; Edwards, 1996; Karasek, 1979; Massimini &
Carli, 1988). Relevant to the activity examined in the present study, Csikszentmihalyi
and Csikszentmihalyi (1988, p. 261) have argued that a beginning piano player will
see learning the keys corresponding to the various notes as challenging, and might
feel in ow simply by running the scales on the keyboard. As soon as the player feels
condent with the scales, however, new challenges need to be found or he or she will
get bored.
Applied to the work situation, this means that employees are more likely to experience ow when their job demands or challenges match their professional skills.
There is balance when, for example, experienced music teachers succeed in teaching

A.B. Bakker / Journal of Vocational Behavior 66 (2005) 2644

29

jazz music to inexperienced students who only learned to play rudimentary and specic pieces of classical music. Such a goal will particularly be reached when employees have certain professional skills (cf. the Person-Environment t model; Edwards,
1996), or have sucient resources in their worksuch as support from colleagues,
performance feedback, good material, and autonomy (Bakker, Demerouti, De Boer,
& Schaufeli, 2003a; Demerouti, Bakker, Nachreiner, & Schaufeli, 2001; Karasek,
1979). Research has indeed shown that such job resources make a positive contribution to the motivation, commitment, and performance of individuals (De Jonge,
1995; Hackman & Oldham, 1980; Loher, Noe, Moeller, & Fitzgerald, 1985; Orpen,
1979; Schaufeli & Bakker, in press), and may buer the undesirable inuence of job
demands on stress reactions (Karasek, 1979; Van Vegchel, De Jonge, Bakker, &
Schaufeli, 2002). Possibilities for self-growth, for example, will enable employees
to cope better with the demands of their work, and feedback about performance
makes clear to what extent work-related goals have been reached (Locke, 1968).
On the basis of these previous ndings, the following hypothesis is formulated (see
also Fig. 1):
Hypothesis 1. Music teachers job resources, including autonomy, performance
feedback, social support from colleagues, and supervisory coaching, have a positive
inuence on the balance between their challenges and skills. This balance, in turn,
contributes positively to the experience of ow (absorption, work enjoyment, and
intrinsic work motivation).

4. Is ow contagious?
Several eld and experimental studies have shown that positive and negative
emotions can crossover from one person to another (for an overview, see Hateld,

Fig. 1. The Flow Model.

30

A.B. Bakker / Journal of Vocational Behavior 66 (2005) 2644

Cacioppo, & Rapson, 1994). For example, Field, Woodson, Cohen, Garcia, and
Greenberg (1982) showed that the facial expression of a newborn child resembles
the facial expression of the mother. Likewise, the emotions of mothers can be inuenced by the facial expressions of their newborn children (Frodi et al., 1978). Campos and Sternberg (1981) have called this social referencing. This phenomenon has
also been described as emotional contagion, dened as: The tendency to automatically mimic and synchronize facial expressions, vocalizations, postures and movements with those of another person and, consequently, to converge emotionally
(Hateld et al., 1994, p. 5). An experiment of Lanzetta, Sullivan, Masters, and
McHugo (1985), in which individuals were lmed and questioned while they were
watching a speech of President Reagan, showed that supporters shared his enjoyment when he was telling something happy, and reported tension when he was telling
something fearful. Although opponents reported negative feelings during the whole
speech, their facial expressions were congruent with those of Reagan. In addition,
assessments of their galvanic skin responses showed that both supporters and opponents were more relaxed during the happy messages than during the disturbing announcements.
Recent organizational studies have shown that employee well-being may take
contagious forms as well. For example, Bakker and Schaufeli (2000) showed that
teachers, general practitioners (Bakker, Schaufeli, Sixma, & Bosveld, 2001), and employees working for an insurance company (Bakker, Demerouti, & Schaufeli, 2003)
caught the burnout symptoms of their colleagues, particularly when they were susceptible to the emotions of others (an individual dierence variable; cf. Sti, Dillard,
Somera, Kim, & Sleight, 1988), and when they frequently discussed work-related
stress. In addition, in a series of crossover studies among couples, (Westman,
2001; Westman & Etzion, 1995; Westman & Vinokur, 1998) has shown that burnout
and depression may transfer from one spouse to another. Taken together, these ndings indicate that our emotions are inuenced by those of others whether we like the
others or not, and whether we want to or not. Note that the evidence found in eld
research for crossover relates solely to negative emotions (e.g., strain, burnout, and
depression). There is little, if any, evidence for positive crossover in the workplace.
However, since stressful demands or a bad day at work have a negative impact on
colleagues well-being, positive experiences at work may also crossover and have a
positive eect on others well-being (cf. Westman, 2001).
It is still unclear which dierent mechanisms play a role in the process of crossover
or emotional contagion. The perception and (unconsciously) adoption of facial expressions seems important (e.g., Adelmann & Zajonc, 1989). For intonation and
speed of speech it has been shown that partners mutually inuence and adjust to
each other, and that the tone height and speed of the speaker inuences the way
he or she feels (Hateld et al., 1994). Thus, based on emotional contagion theory,
we may expect that ow experiences of music teachers have a positive inuence on
those of their students. During the teaching, several ongoing interactive (verbal as
well as non-verbal) processes take place between the teacher and his or her students.
In addition to seeing each other and talking to each other, the teacher plays an important role in designing the lessons. The mood of the teacher thus inuences many

A.B. Bakker / Journal of Vocational Behavior 66 (2005) 2644

31

factors to which students are exposed during the lesson, and in this way, the processes of emotional contagion may take eect. In addition, the ow experience of
a teacher not only includes a component of work enjoyment, which may inuence
the student, it also contains the components of intrinsic work motivation and absorption. This may determine the quality of the lesson and, indirectly, the enjoyment
of the student. Therefore, the second hypothesis is:
Hypothesis 2. There exists a positive relationship between the experience of ow
(absorption, work enjoyment, and intrinsic work motivation) of music teachers and
the ow experiences of their students.

5. Method
5.1. Procedure and participants
In consultation with the Royal Dutch Society for Music Artists (Koninklijke Nederlandse Toonkunstenaars Bond), 28 music schools were approached, of which 16
participated (57% of the schools). The 405 music teachers who were employed at
these schools received questionnaires in their mailboxes at work. In addition to questions about demographics, the questionnaire included items to assess job resources
and the balance between challenges and skills, as well as items about work-related
ow experiences. The accompanying letter explained that the goal of the study
was to examine teacher well-being. In addition, all teachers were asked to randomly
distribute a short ow questionnaire (including the three dimensions of absorption,
enjoyment, and intrinsic motivation) to four of their students during one of the lessons. The items were somewhat adjusted for the students, by referring to ow experiences when playing their musical instrument. In total, 178 teachers (44%) and 605
students lled out the questionnaire. Reasons not to ll out the questionnaire were:
I am too busy, We already have to ll out too many forms, and Participation
is not anonymous if I ll out my age and instrument. For 67% of all teachers, 4 students lled out the ow questionnaire, for 80% of all teachers at least 3 students lled
out the questionnaire, and for 92% of all teachers at least 2 students lled out the
questionnaire (for all teachers at least 1 student). Teachers and students turned in
the questionnaires separately at the reception desk of their music school.
The sample included 75 men (42%) and 103 women (58%). On average, they were
teaching 12 h (SD 7:1 h) per week at their music school and organizational tenure
was on average 13 years (SD 10:2 years). The mean age of the participants was 41
years (SD 9:8). Of all teachers, 40% performed other activities as well, such as providing private lessons and playing in an ensemble or professional orchestra (or both).
In total, 36% only taught individual students. The remainder taught both groups and
individuals. The average age of the students was 19 years (SD 13:2). On average,
they had received music lessons for ve years and generally studied between 2 and 4 h
per week. The teaching concerned several instruments, particularly piano (17%), ute
(12%), guitar (12%), violin (10%), and keyboard (9%). Regarding genre, 57% of the

32

A.B. Bakker / Journal of Vocational Behavior 66 (2005) 2644

teachers taught only classical music, 27% classic and pop music, and 16% only pop
music. In addition, 20% of the teachers included jazz music in the curriculum, always
in combination with classical music or pop music. Preliminary analyses showed that
none of the demographics was related to the model variables. Therefore, these factors will not be included in the following analyses.
5.2. Measures
Job resources. Four job resources were included in the questionnaire. Autonomy
was measured with a short scale developed by Bakker, Demerouti, Taris, Schaufeli,
and Schreurs (2003c). Autonomy or job control is considered a job resource since
many studies have shown that this work characteristic enables employees to cope
with job demands (Van der Doef & Maes, 1999), and has motivational potential
(Fried & Ferris, 1987; Hackman & Oldham, 1980). The scale includes three items referring to decision authority (i.e., freedom of action in accomplishing the formal
work task). A sample item is: Can you decide yourself how you execute your
work? Items are scored on a ve-point Likert scale (1 never, 5 always). Social
support was also assessed with a three-item scale developed by Bakker et al. (2003c).
A sample item is: Can you ask your colleagues for help if necessary? (1 never, 5
always). Coaching by the supervisor was measured using a Dutch adaptation of
Graen and Uhl-Beins (1991) Leader-Member exchange scale (Le Blanc, 1994).
The scale includes ve items, such as: My supervisor uses his/her inuence to help
me solve my problems at work (1 never, 5 always). Performance feedback was
assessed with a three-item scale developed by Bakker et al. (2003c). A sample item
is: I receive sucient information about the goal of my work (1 never, 5 always). All responses were coded such that higher scores referred to more job resources.
The Balance between challenges and skills was assessed with a ve-item scale,
based on French, Caplan, and Harrison (1982). Sample items are: I am well able
to meet the demands of my work, and I have sucient skills to carry out my work
tasks properly.
Flow was assessed with a recently developed instrument named the WOrk-reLated
Flow scale (WOLF for short; Bakker, 2001). The WOLF includes 13 items measuring absorption (4 items), work enjoyment (4 items) and intrinsic work motivation (5
items). Examples are: When I am working, I forget everything else around me (absorption), When I am working very intensely, I feel happy (work enjoyment), and
I get my motivation from the work itself, and not from the rewards for it (intrinsic
work motivation). The participants were asked to indicate how often they had each of
the experiences during the preceding week (0 never, 1 almost never, 2 sometimes, 3 regularly, 4 often, 5 very often, 6 always). Seven previous studies among a total of 1346 employees from dierent organizations and occupations
have shown that the three factors can be empirically distinguished, and have good
reliabilities (Bakker, 2001). Cronbachs a-coecients for absorption, work enjoyment and intrinsic work motivation ranged from .75 to .86, .88 to .96, and .63 to
.82, respectively. In addition, a longitudinal study among 248 consultants who lled

A.B. Bakker / Journal of Vocational Behavior 66 (2005) 2644

33

out the WOLF twice with 6 weeks between both waves showed that the testretest
reliabilities of the subscales are satisfactory (Bakker, 2001). The testretest correlations were .74, .77, and .71 for absorption, work enjoyment, and intrinsic work motivation, respectively. To determine ow among students, the items were slightly
adjusted. All items that refer to work were changed by referring to playing music,
for example: When I am playing music, I forget everything else around me. The
scores on each of the three ow dimensions were calculated for groups of students
by using the average of all students for each teacher.
5.3. Analyses
To test the two hypotheses simultaneously, structural equation modeling (SEM)
analyses were carried out with the AMOS software package (Arbuckle, 1997). In order to test the t between the model and the data, the traditional v2 value, the goodness-of-t index (GFI) and the root mean square error of approximation (RMSEA)
were calculated. As a rule of thumb, a GFI P .90 and a RMSEA P .08 indicate a
reasonable t between the model and the data (Browne & Cudeck, 1993). Because
these indices are dependent on sample size, as recommended by Marsh, Balla, and
Hau (1996)the non-normed t index (NNFI), the incremental t index (IFI),
and the comparative t index (CFI) were also examined. These indices should have
values of .90 or higher (Hoyle, 1995).
The model, as displayed in Fig. 1, consists of hypothetical constructs or latent
variables that are all estimated by two or more manifest variables that are directly
observed (i.e., the scales introduced before). The latent exogenous factor job resources was operationalized by four observed, manifest variables, namely social support, supervisory coaching, autonomy, and performance feedback. In addition, the
structural model includes two latent mediating variables: (1) balance between challenges and skills, and (2) ow among music teachers. The scale for assessing balance
was split in two reliable halves (cf. Bakker et al., 2000) that served as the indicators
of the latent variable balance. The rst subscale includes two items and has a reliability of .76 (r :61, p < :001). The second subscale includes three items and has a
reliability of .79. The latent factor ow among music teachers was indicated by absorption, work enjoyment, and intrinsic work motivation. Finally, the three respective ow scales indicated the latent endogenous factor ow among students.

6. Results
6.1. Descriptive statistics
Table 1 shows the means, standard deviations, reliability coecients, and correlations between all study variables. As can be seen, all constructs that were assessed
demonstrated good internal consistencies. In addition, Table 1 shows that the four
job resources correlate signicantly with each other, and with teachers ow experiences. The intercorrelations between the three ow components are relatively high,

34

Variable
1 Social support by
colleagues
2 Supervisory coaching
3 Autonomy
4 Performance feedback
5 Balance (indicator 1)
6 Balance (indicator 2)
7 Absorption teacher
8 Work enjoyment teacher
9 Intrinsic work motivation
teacher
10 Absorption Students
11 Enjoyment students
12 Intrinsic motivation
students
*

p < :05.
p < :01.

**

SD

3.37

1.07

(.86)

3.17
3.70
2.89
4.28
4.20
3.78
4.67
3.98

.98
.74
1.08
.62
.64
1.11
.93
1.02

.39


.28


.33


.20


.12
).10
.23


.13

3.73
4.76
4.72

.83
.56
.49

.30


.18

.07

(.90)
.39


.33


.11
.01
).08
.07
).02

(.75)
.30


.08
.07
.03
.18

.18

(.77)
.18

.22


).12
.18

.07

(.76)
.75


.17

.28


.21

(.79)
.20


.34


.24

(.76)
.44


.50

(.88)
.66

(.71)

.05
.03
).02

).02
.09
.14

.15

.10
.10

.06
.15

.11

.09
.05
.09

).01
.09
).04

).06
.10
.09

.18

.27


.27

10

(.91)
.62


.40

11

12

(.90)
.69

(.91)

A.B. Bakker / Journal of Vocational Behavior 66 (2005) 2644

Table 1
Means, standard deviations, reliability coecients (Cronbachs a; on the diagonal), and correlations between the variables, N 178 music teachers and 605
students

A.B. Bakker / Journal of Vocational Behavior 66 (2005) 2644

35

for both the teachers and the students. Furthermore, teachers ow experiences
partly correlate with those of the students. Another noticeable nding is that social
support reported by teachers correlates signicantly with students levels of absorption and enjoyment when playing music.
6.2. Test of the ow model
According to Hypothesis 1, job resources have a positive relationship with the
balance between challenges and skills, which, in turn, contributes to explaining variance in music teachers experiences of ow (absorption, work enjoyment, and intrinsic work motivation). In addition, Hypothesis 2 states that there is a positive
relationship between teachers ow and that of their students (contagion hypothesis;
see also Fig. 1). To test these hypotheses simultaneously, SEM analyses were carried
out with the AMOS software package (Arbuckle, 1997).
The rst row in Table 2 shows that the hypothesized model ts reasonably well to
the data. Only the NNFI is, at .87, somewhat lower than the criterion value of .90.
The GFI, CFI, and IFI are all .90, andtogether with a RMSEA of .08this indicates an acceptable t between the model and the data. In addition, all indicators
had signicant loadings on the intended factors. For job resources, these factor loadings ranged from .54 to .63. The factor loadings for absorption, (work) enjoyment
and intrinsic (work) motivation were .59, .80, and .74, respectively, for ow among
teachers, and .63, .95, and .71 for ow among students.
To test the alternative hypothesis that job resources also have a direct relationship
with ow among teachers and ow among students, both paths were included in the
model. The results (see Table 2) showed that this alternative model (M2) does not t
better to the data than the proposed hypothetical model, Delta v2 2 1:74, n.s.
Moreover, both path coecients had the same, non-signicant value of .10
(t0 s < 1). In a third model (M3), the path from balance to ow among students
was added to M1. This did also not result in a better t of the model to the data (Delta v2 1 :27, n.s.), and the coecient of the additional path proved to be non-sig-

Table 2
Results of SEM-analyses: Fit indices of the ow model and the alternative models, standardized maximum
likelihood estimates, N 178 music teachers and 605 students
Model

v2

df

GFI

RMSEA

NNFI

CFI

IFI

M1.
M2.
M3.
M0.

113.50
111.76
113.23
673.94

51
49
50
66

.90
.91
.90
.59

.08
.08
.08
.23

.87
.87
.87

.90
.90
.90

.90
.90
.90

Flow Model
Alternative Model
Alternative Model
Null Model

Note: v2 , chi-square; df , degrees of freedom; GFI, goodness-of-t index; RMSEA, root mean square
error of approximation; NNFI, non-normed t index; CFI, comparative t index; IFI, incremental t
index; M2, alternative model, including the path from job resources to ow among teachers, and from job
resources to ow among students; and M3, alternative model, including the path from balance to ow
among students.

36

A.B. Bakker / Journal of Vocational Behavior 66 (2005) 2644

Fig. 2. The Flow Model: Standardized solution (maximum likelihood estimates), N 178 teachers/605
students.

nicant (b :17, t 1:52). These results of the alternative models oer additional
support for the hypothetical model.
The results of the nal ow model, which is identical to the hypothetical model,
are summarized in Fig. 2. The model explains 5% of the variance in balance between
challenges and skills, 16% of the variance in ow among teachers, and 6% of the variance in ow among students.
6.3. Flow as a peak experience
In all previous analyses, ow was conceptualized as a phenomenon with three underlying dimensions (absorption, (work) enjoyment, and intrinsic (work) motivation), and it was assumed that individuals could score low or high on each of
them. In other words, the three ow dimensions were operationalized as continuous
variables, and the latent factor ow was indicated by each of these variables in the
SEM-analyses. However, one could argue that the peak experience of ow is an
all or nothing phenomenon: an experience that one either has or does not have (dichotomous variable). To examine whether this alternative operationalization of ow
would lead to dierent results, an additional SEM-analysis was performed. Because
externally validated criteria for cut-off scores on each dimension do not exist, the following rule of thumb was used: teachers and students have a peak experience if they
score higher than or equal to the 75e percentile of each of the three ow dimensions.
According to this criterion, 19 teachers (10.7%) and 20 groups of students (11.2%)
reported experiences of ow (and were assigned the number 1 in the analyses).
The ow model was tested again, but now ow was included in the model as an observed, manifest variable (participants scored 0 or 1). This model also turned out to
t the data well, v2 19 26:31, p :12, GFI :96, RMSEA :05, NNFI :96,
CFI :97, IFI :97. In this model, the relationship between balance and ow

A.B. Bakker / Journal of Vocational Behavior 66 (2005) 2644

37

among teachers was 17. (CR 1:98, p < :05), and the relationship between ow
among teachers and ow among students was also .17 (CR 2:23, p < :05). Taken
together, these ndings are consistent with those in Fig. 2, although the relationships
in the model are less strong.
It should be noted that maximum likelihood (ML) estimation assumes that variables are continuous instead of dichotomous. Unfortunately, Asymptotically Distribution Free (ADF) estimation cannot be used to circumvent this problem in the
present study, because of the relative small sample size (cf. Hoogland & Boomsma,
1997). I therefore checked the appropriateness of the ndings by comparing the ML
estimations with the results of logistic regression analyses (cf. Jones, 1998). Logistic
regression is similar to linear regression but is suited to models where the dependent
variable is dichotomous. The results of this analysis conrmed that the balance between challenges and skills is signicantly related to ow among music teachers
(B 1:15, p < :05), and that ow among music teachers is signicantly related to
ow among students (B 1:23, p < :05).

7. Discussion
The central aim of this study was to answer two questions, namely whether job
resources are possible antecedents of ow experiences among music teachers, and
to what extent ow may crossover from teachers to their students. To answer both
questions, ow at work was rst dened as a short-term peak experience that is characterized by absorption, work enjoyment, and intrinsic work motivation (cf. Bakker,
2001; Csikszentmihalyi, 1997; Ellis et al., 1994). Individuals who experience ow forget everything else around them, evaluate the quality of their working life very positively, and perform certain work-related activities with the aim to experience
enjoyment and satisfaction that is inherent to these activities. In addition, on the basis of previous research, it was predicted that ow is most likely when individuals
perceive a balance between the challenges of a situation and their own skills to cope
with these challenges (e.g., Clarke & Haworth, 1994; Csikszentmihalyi, 1990; Ellis
et al., 1994; Massimini & Carli, 1988). Resources in the working environment would
facilitate this balance, because job characteristics such as autonomy, performance
feedback, social support, and supervisory coaching can foster personal growth and
the realization of goals (cf. Bakker et al., 2003c; Demerouti et al., 2001). Therefore,
the rst hypothesis was that job resources would have a positive relationship with
music teachers balance between their challenges and skills, which, in turn, would
contribute to explaining ow (absorption, work enjoyment and intrinsic work motivation).
The results of SEM-analyses supported this hypothesis. More specically, the
ndings showed that job resourcesa combination of autonomy, performance
feedback, social support from colleagues and supervisory coachinghad a positive
relationship with the balance between challenges and skills, and that this balance, in
turn, had predictive value for the frequency of ow among music teachers. These
ndings conrm and expand Csikszentmihalyis (1997, 1990) ow theory. They

38

A.B. Bakker / Journal of Vocational Behavior 66 (2005) 2644

emphasize the importance of the t between the skills of the person and the challenges he or she is exposed to at work (cf. French et al., 1982). In addition, the results
show that besides the facilitating role of information about the goals and results of
ones work (cf. Csikszentmihalyi), social support from colleagues, supervisory coaching and autonomy may also contribute to such a balance.
However, there is another issue at stake here. In terms of balance, one may specify
that it involves equilibrium between job demands (challenges) and skills, but more
generally, one may argue that balance involves equilibrium between job demands
and all accessible resources (personal resources, including e.g. skills, self-ecacy,
and self-esteem; and job resources, including performance feedback, autonomy, social support, and coaching). Resources in excess of demands may yield increased
boredom, and demands in excess of resources may yield increased distress. This
means that if job demands are held constant, resources and the likelihood of ow will
take on a quadratic function (cf. Warr, 1987). As resources increase, the likelihood of
ow will increase, presumably mitigating stress along the way. This positive relationship should hold until equilibrium between job demands and resources is achieved.
As resources exceed the xed job demands, then the likelihood of a ow state may
decline because boredom will become the more likely state. Some evidence for such
a pattern of responses has indeed been found in previous job stress studies with the
demand-control model (De Jonge & Kompier, 1997) and the job demandsresources model (Bakker, Demerouti, & Euwema, 2003b).
The second hypothesis was that there exists a positive relationship between music teachers ow (absorption, work enjoyment, and intrinsic work motivation)
and the experience of ow among their students. This hypothesis was also conrmed by the ndings; the more ow experiences music teachers reported, the
higher the frequency of comparable experiences among their students. This nding
is in line with emotional contagion theory (Hateld et al., 1994), and is one of the
rst demonstrations in eld research that positive emotions may crossover from
one person to another (Westman, 2001). The correlational analysis suggested that,
in particular, teachers intrinsic work motivation was related to ow experienced by
students. This ow dimension showed the strongest positive relationships with students intrinsic motivation, and was also related to their absorption and enjoyment. In addition, music teachers and students absorption and enjoyment also
showed positive relationships, and enjoyment had the highest loading on the latent variable ow in the SEM-analyses. Thus, these results suggest that students
caught the ow experience of their teachers in a partly conscious and partly unconscious way. The crossover process may include the automatic imitation of a
cheerful and happy teacher, but also the more conscious crossover of teachers
dedication to their work (cf. Hateld et al., 1994). In addition, motivated teachers
probably put more eort and energy in the search for nice and suitable music for
the students; they presumably have a more positive attitude toward work that
motivates their students to concentrate during the music lesson and to perform
well.
In previous studies, researchers have applied various methods to assess ow.
Csikszentmihalyi (1990) usually applies the experience sampling method

A.B. Bakker / Journal of Vocational Behavior 66 (2005) 2644

39

(Csikszentmihalyi & Larson, 1992; Kimiecik & Stein, 1992). Participants are
requested to react several times a day to a short questionnaire (experience sampling
form) when they receive a random signal from an electronic device during a certain
period of time (usually one week). Jackson and Marsh (1996) have argued that more
research is needed to examine the reliability and validity of this method in eld research, and ask for attention to the practical problems that are inherent in this research strategy (see Kimiecik & Stein, 1992). The results of the present study
suggest that simple questionnaires may oer a reasonable alternative. Note, however, that Csikszentmihalyi (1990), although primarily assessing the three dimensions
of absorption, enjoyment, and intrinsic motivation, has described several other elements of ow, including the balance between challenges and skills, a clear goal, feedback, concentration and focus, control, loss of self-consciousness, transformation of
time, and an autotelic experience (the activity becomes a goal in itself). According
to Csikszentmihalyi, the frequency of ow will increase with an increase in each of
these eight factors. More research is needed to nd out whether the three elements
used in the current study suciently measure the experience of ow.
In the present study, the three ow dimensions were each assessed on a continuum, ranging from never to always. The question is, therefore, to what extent
we really examined peak experiences among teachers and their students. One could
argue that individuals only experienced ow if they scored high on each of the three
ow factors (for example, P the 75e percentile). The results of additional analyses
showed that approximately 11% of all teachers and students met this criterion, and
that the relationships in the ow model were comparable with those between continuous variables (although weaker). The disadvantage of this approach is that nuances
in the experience of ow are lost, since one does not distinguish between the three
underlying dimensions (absorption, (work) enjoyment and intrinsic (work) motivation). The advantage of a combined measure is that it may be very useful for organizational practice. Interestingly, the conceptualization of ow as a combined
measure is comparable with the approach of job burnout by Schaufeli and Van Dierendonck (2000), who dene the risk of burnout as a combination of relatively high
scores on the three dimensions measuring the syndrome, namely exhaustion, cynicism, and reduced professional ecacy.
7.1. Study limitations
A limitation of the present study is its cross-sectional character. Thus, the current ndings, although framed in terms of cause and eect, did not demonstrate
causality but merely relationships. Using existing theories, certain causeeect relationships were assumed, but the existence of reversed causal eects cannot be
excluded. For example, it is conceivable that teachers also experience more ow
when their students have more of such peak experiences. Therefore, it is desirable
that future studies examine such reciprocal eects. However, a strong point of this
study is that two sources of information were used, namely music teachers and
their students. Common method variance is a non-issue for the contagion hypothesis in this study.

40

A.B. Bakker / Journal of Vocational Behavior 66 (2005) 2644

A second limitation is that only teachers self-reports were used in the examination of relationships between job resources, balance and ow. Here the problem of common method variance may have played a role. Because the study was
conducted at 16 dierent schools, we could examine to what extent the ow experience per school was a function of the resources available at that school. Additional analyses showed that the scores on the ow dimensions absorption,
work enjoyment, and intrinsic work motivation were highest at those schools
where many resources were available (particularly supervisory coaching and autonomy). These ndings indicate that the subjectively reported job characteristics
were anchored in the objective working situation. The results thus oer specic
starting-points for interventions aimed at mobilizing job resources and the promotion of ow experiences at music schools.
Third, it should be noted that the model explained only a limited amount of
variance (16% of the variance in ow of teachers, and 6% of the variance of
the ow among students). This can partly be explained by the use of dierent
sources of information, which leads to an underestimation of the strength of relationships (cf. Frese & Zapf, 1988). It can be expected that both sources of information (in this study teachers and students) have their own unique causes of
(statistically independent) error variance. If one uses only one method, the sources
of error variance are the same. The consequence of this is inated correlations. In
addition, it should be noted that our strategy of analysis reduced the variance in
students ow experiences. More specically, for each teacher, students scale
scores on each of the ow dimensions were summed for groups of students. This
implies that the variance in the student responses was restricted. Future studies
with a similar design may circumvent this drawback by using multi-level analyses
techniques.
Fourth, although teachers were asked to randomly distribute short ow questionnaires among their students, it is unclear to what extent they complied with
this request. It cannot be excluded that the teachers were selective. On the other
hand, most teachers (67%) found four students who were willing to indicate the extent to which they experienced absorption, enjoyment and intrinsic motivation during the lessons. A limitation that we cannot overcome is that the average scores of
the students on the three ow dimensions were, in some cases, based on less than
four students.
Finally, regarding the crossover of ow, this study is limited to the context of music teachers and their students. Since the contagion hypothesis was conrmed regarding ow, it would be interesting and relevant to examine this phenomenon in other
teacherstudent relationships, such as in non-traditional teacherstudent relationships (i.e. sport teachers and their students). In addition, it would be interesting to
nd out whether ow may also crossover from employees to their colleagues or partners. Several recent studies have shown that stress and strain can crossover at work
(Bakker & Schaufeli, 2000; Bakker et al., 2003) and from work to private life (e.g.,
Westman, 2001). From a practical and theoretical point of view it would be valuable
to carry out more organizational studies to examine whether positive experiences, including ow, can crossover to others.

A.B. Bakker / Journal of Vocational Behavior 66 (2005) 2644

41

7.2. Practical implications

It can be concluded that it is important for music teachers to have sucient
resources available in their work. Autonomy, social support from colleagues,
supervisory coaching, and performance feedback are examples of such resources,
but one could also consider skill variety, possibilities for self-growth, participation
in decision making, and high quality communication (see also Bakker et al.,
2003a, 2003b, 2003c). After completion of the present study, the participating
schools received feedback about their most important job resources. Using histograms, it was explained how teachers at each school scored relative to the research group as a whole and in comparison with an external benchmark,
including information that has been collected in recent years in many dierent
companies (e.g., Bakker et al., 2003b, 2003c). On the basis of this information,
it is possible to optimize the working environment. If, for example, one schools
scores are relatively low on supervisory coaching and the ow scores of the teachers and students are inferior to those of teachers in other schools, it may be desirable to oer additional training to supervisors regarding leadership styles.
Feedback could be mobilized as a resource by making job evaluations, and by
regularly informing music teachers about the results of their work. Since we know
that music teachers experiences of ow coincide with those of their students, such
resources seem not only important for teachers themselves, but also for the happiness and motivation of their students.

References
Adelmann, P. K., & Zajonc, R. B. (1989). Facial reference and the experience of emotion. Annual Review
of Psychology, 40, 249280.
Arbuckle, J. L. (1997). Amos users guide. Version 3.6. Chicago: Smallwaters.
Bakker, A. B. (2001). Vragenlijst voor het meten van werkgerelateerde ow: De WOLF [Questionnaire for
the assessment of work-related ow: The WOLF]. Utrecht University, Utrecht, The Netherlands:
Department of Social and Organizational Psychology.
Bakker, A. B., Demerouti, E., De Boer, E., & Schaufeli, W. B. (2003a). Job demands and job resources as
predictors of absence duration and frequency. Journal of Vocational Behavior, 62, 341356.
Bakker, A. B., Demerouti, E., & Euwema, M. (2003). How job resources buer the impact of job demands on
burnout. Manuscript submitted for publication.
Bakker, A.B., Demerouti, E., & Schaufeli, W. B. (in press). The socially induced burnout model. In F.
Columbus (Ed.), Advances in Psychology Research. New York: Nova Science Publishers.
Bakker, A. B., Demerouti, E., Taris, T., Schaufeli, W. B., & Schreurs, P. (2003c). A multi-group analysis
of the job demands - resources model in four homecare organizations. International Journal of Stress
Management, 10, 1638.
Bakker, A. B., & Schaufeli, W. B. (2000). Burnout contagion processes among teachers. Journal of Applied
Social Psychology, 30, 121.
Bakker, A. B., Schaufeli, W. B., Sixma, H., & Bosveld, W. (2001). Burnout contagion among general
practitioners. Journal of Social and Clinical Psychology, 20, 8298.
Browne, M. W., & Cudeck, R. (1993). Alternative ways of assessing model t. In K. A. Bollen & J. S. Long
(Eds.), Testing structural equation models. Sage Publications.
Campos, J. J., & Sternberg, C. R. (1981). Perception, appraisal and emotion: The onset of social
referencing. In M. Lamb & L. Sherrod (Eds.), Emotion (pp. 184). Monterey, CA: Brooks/Cole.

42

A.B. Bakker / Journal of Vocational Behavior 66 (2005) 2644

Catley, D., & Duda, J. L. (1997). Psychological antecedents of the frequency and intensity of ow in
golfers. International Journal of Sport Psychology, 28, 309322.
Clarke, S. G., & Haworth, J. T. (1994). Flow experience in the daily lives of sixth-form college students.
British Journal of Psychology, 85, 511523.
Csikszentmihalyi, M. (1990). Flow: The psychology of optimal experience. New York: HarperCollins.
Csikszentmihalyi, M. (1997). Finding ow: The psychology of engagement with everyday life. New York:
HarperCollins.
Csikszentmihalyi, M.& Csikszentmihalyi, I. S. (Eds.). (1988). Optimal experience: Psychological studies of
ow in consciousness. New York: Cambridge University Press.
Csikszentmihalyi, M., & Larson, R. (1992). Validity and reliability of the experience sampling method. In
M. de Vries (Ed.), The experience of psychopathology. New York: Cambridge University Press.
Csikszentmihalyi, M., & LeFevre, J. (1989). Optimal experience in work and leisure. Journal of Personality
and Social Psychology, 56, 815822.
Csikszentmihalyi, M., Rathunde, K., & Whalen, S. (1993). Talented teenagers: The roots of success and
failure. New York: Cambridge University Press.
Deci, E. L., & Ryan, R. M. (1985). Intrinsic motivation and self-determination in human behaviour. New
York: Plenum Press.
De Jonge, J. (1995). Job autonomy, well-being and health. Unpublished Doctoral Dissertation. Maastricht,
The Netherlands: University of Maastricht.
De Jonge, J., & Kompier, M. A. J. (1997). A critical examination of the Demand-Control-Support
Model from a work psychological perspective. International Journal of Stress Management, 4,
235258.
Demerouti, E., Bakker, A. B., Nachreiner, F., & Schaufeli, W. B. (2001). The job demandsresources
model of burnout. Journal of Applied Psychology, 86, 499512.
Diener, E. (2000). Subjective well-being, the science of happiness and a proposal for a national index.
American Psychologist, 55, 3443.
Diener, E., & Diener, C. (1996). Most people are happy. Psychological Science, 7, 181185.
Edwards, J. R. (1996). An examination of competing versions of the person-environment t approach to
stress. Academy of Management Journal, 39, 292339.
Ellis, G. D., Voelkl, J. E., & Morris, C. (1994). Measurement and analyses issues with explanation of
variance in daily experience using the ow model. Journal of Leisure Research, 26, 256337.
Field, T., Woodson, R., Cohen, D., Garcia, R., & Greenberg, R. (1982). Discrimination and
imitation of facial expressions by term and preterm neonates. Infant Behaviour Development, 6,
485490.
French, J. R. P., Jr., Caplan, R. D., & Harrison, R. V. (1982). The mechanisms of job stress and strain. New
York: Wiley.
Frese, M., & Zapf, D. (1988). Methodological issues in the study of work stress: Objective vs. subjective
measurement of work stress and the question of longitudinal studies. In C. L. Cooper & R. Payne
(Eds.), Causes, coping and consequences of stress at work (pp. 375411). Chichester: Wiley.
Fried, Y., & Ferris, G. R. (1987). The validity of the job characteristics model: A review and meta-analysis.
Personnel Psychology, 40, 287322.
Frodi, A. M., Lamb, M. E., Leavitt, L. A., Donovan, W. L., Ne, C., & Sherry, D. (1978). Fathers and
mothers responses to the faces and cries of normal and premature infants. Developmental Psychology,
14, 490498.
Ghani, J. A., & Deshpande, S. P. (1994). Task characteristics and the experience of optimal ow in humancomputer interaction. The Journal of Psychology, 128, 381391.
Graen, G. B., & Uhl-Bein, M. (1991). The transformation into professionals into self-managing and
partially self-designing contributors: Toward a theory of leadership making. Journal of management
Systems, 3, 2539.
Hackman, J. R., & Oldham, G. R. (1980). Work redesign. Reading, MA: Addison Wesley.
Harackiewicz, J. M., & Elliot, A. J. (1998). The joint eects of target and purpose goals on intrinsic
motivation: A mediational analysis. Personality and Social Psychology Bulletin, 24, 657689.
Hateld, E., Cacioppo, J. T., & Rapson, R. L. (1994). Emotional contagion. New York: Cambridge
University Press.

A.B. Bakker / Journal of Vocational Behavior 66 (2005) 2644

43

Hoogland, J., & Boomsma, A. (1997). Robustness studies in covariance structure modeling: An overview
and meta analysis. Sociological Methods and Research, 26, 329367.
Hoyle, R. H. (1995). The structural equation modeling approach: Basic concepts and fundamental issues.
In R. H. Hoyle (Ed.), Structural equation modeling, concepts, issues, and applications (pp. 115).
Thousand Oaks, CA: Sage.
Jackson, S. A., & Marsh, H. W. (1996). Development and validation of a scale to measure optimal
experience: The ow state scale. Journal of Sport and Exercise Psychology, 18, 1735.
Jones, F. F. (1998). Pay procedures and voluntary turnover: Does procedural justice matter? Psychological
Reports, 83, 475482.
Karasek, R. A. (1979). Job demands, job decision latitude, and mental strain: Implications for job design.
Administrative Science Quarterly, 24, 285308.
Kimiecik, J. C., & Stein, G. L. (1992). Examining ow experiences in sport contexts: Conceptual issues and
methodological concerns. Journal of Applied Sport Psychology, 4, 144160.
Klausmeier, F. (1978). Die Lust sich musikalisch aus zu dr
ucken [The passion to express oneself with music].
Hamburg: Rowohlt Verlag.
Kowal, J., & Fortier, M. S. (1999). Motivational determinants of ow: Contributions from selfdetermination theory. The Journal of Social Psychology, 139, 355368.
Lanzetta, J. T., Sullivan, D. G., Masters, R. D., & McHugo, G. J. (1985). Viewers emotional and
cognitive responses to televised images of political leaders. In S. Kraus & R. Perlo (Eds.), Mass media
and political thought (pp. 85116). Beverly Hills, CA: Sage.
Larson, R., & Richards, M. H. (1994). Divergent realities: The emotional lives of mothers, fathers, and
adolescents. New York: Basic Books.
Le Blanc, P. (1994). De steun van de leiding: Een onderzoek naar het Leader Member Exchange model in de
verpleging [Leaders support: A study of the Leader Member Exchange model among nurses].
Amsterdam: Thesis Publishers.
Locke, E. A. (1968). Eects of knowledge of results, feedback in relation to standards, and goals on
reaction-time performance. American Journal of Psychology, 81, 566574.
Loher, B. T., Noe, R. A., Moeller, N. L., & Fitzgerald, M. P. (1985). A meta-analysis of the relation of job
characteristics and job satisfaction. Journal of Applied Psychology, 70, 8089.
Lutz, R. J., & Guiry, M. (1994). Intense consumption experiences: Peaks, performances, and ows. Paper
presented at the Winter Marketing Educators Conference, St. Petersburg, FL (February).
Marsh, H. W., Balla, J. R., & Hau, K. T. (1996). An evaluation of incremental t indices: A
clarication of mathematical and empirical properties. In G. A. Marcoulides & R. E. Schumacker
(Eds.), Advanced structural equation modeling, issues and techniques (pp. 315353). Mahwah, NJ:
Lawrence Erlbaum.
Massimini, F., & Carli, M. (1988). The systematic measurement of ow in daily experience. In M.
Csikszentmihalyi & I. S. Csikszentmihalyi (Eds.), Optimal experience: Psychological studies of ow in
consciousness. Cambridge: Cambridge University Press.
Orpen, C. (1979). The eects of job enrichment on employee satisfaction, motivation, involvement, and
performance: A eld experiment. Human Relations, 32, 189217.
Palmer, C. (1989). Music therapy in gerontology: A review and a projection. Music Therapy Perspectives,
6, 5256.
Schaufeli, W. B., & Bakker, A. B. (in press). Job demands, job resources, and their relationship with
burnout and engagement: A multi sample study. Journal of Organizational Behavior.
Schaufeli, W. B., & Van Dierendonck, D. (2000). De UBOS, Utrechtse Burnout Schaal. Handleiding [The
UBOS, Utrecht Burnout Scale. Manual]. Lisse, The Netherlands: Swets Test Services.
Sti, J. B., Dillard, J. P., Somera, L., Kim, H., & Sleight, C. (1988). Empathy, communication, and
prosocial behavior. Communication Monographs, 55, 198213.
Van der Doef, M., & Maes, S. (1999). The Job Demand-Control (-Support) Model and psychological wellbeing: A review of 20 years of empirical research. Work and Stress, 13, 87114.
Van Vegchel, N., De Jonge, J., Bakker, A. B., & Schaufeli, W. B. (2002). Testing global and specic
indicators of rewards in the Eort-Reward Imbalance Model: Does it make any dierence? European
Journal of Work and Organizational Psychology, 11, 403421.
Veenhoven, R. (1984). Conditions of happiness. Dordrecht, The Netherlands: Kluwer.

44

A.B. Bakker / Journal of Vocational Behavior 66 (2005) 2644

Veenhoven, R. (1996). Leefbaarheid van landen. Rede uitgesproken bij aanvaarding van Bijzonder
Hoogleraarschap, Universiteit van Utrecht.
Warr, P. B. (1987). Work, unemployment and mental health. Oxford: Oxford University Press.
Westman, M. (2001). Stress and strain crossover. Human Relations, 54, 717751.
Westman, M., & Etzion, D. (1995). Crossover of stress, strain and resources from one spouse to another.
Journal of Organizational Behavior, 16, 169181.
Westman, M., & Vinokur, A. (1998). Unraveling the relationship of distress levels within couples:
Common stressors, empathic reactions, or crossover via social interactions? Human Relations, 51, 137
156.